“…As possible determinants of programs, loans, and compliance, the initial regressions included a huge number of variables that have been suggested in the literature: the rate of monetary expansion, the overall budget deficit, general government consumption relative to GDP, real GDP growth, GDP per capita, the share of foreign short-term debt in total foreign debt, the total level of outstanding debt, total debt service (in percent of GDP), the rate of inflation, a country's international reserves (in months of imports), the current account balance as a percent of GDP, openness to trade (all World Bank data) and the LIBOR on three months credits to US banks (IMF 2003). I include the following political and social variables: the degree of democracy (Marshall and Jaggers 2000), a measure of political instability (Dreher 2002), 23 fractionalization of the legislature, proportional representation, a dummy for special interest governments, government ideological cohesion and the duration of the political regime (all from Beck et al 2001). All regressions also include a dummy for each individual country; where necessary an AR(1) term is included to correct for serial correlation.…”